or THE SEEDS OF BAERIKGTONIA AND OAREYA. 49 



bined layers. As these authors possessed no ripe seeds, this cha- 

 racter is an inference from a careful comparison of Gsertner's figure 

 and description with the descriptions of Koxburgh and Blume, 

 and perhaps with Eoxburgh's drawing in the India House. As 

 the result of this comparison, they state that they have no doubt 

 that the structure in both genera is identical, and that the sup- 

 posed albumen is part of the embryo, while no real albumen 

 exists. 



In 1839-40 Endlicher (* Genera Plantarum ') ascribes to both 

 genera an exalbuminous embryo united with the cotyledons into a 

 homogeneous fleshy mass, and makes no allusion to the double 

 layer clearly indicated in Gaertner's figure, and described by Rox- 

 burgh and Wight and Amott. 



In 1841 Wight, returning to the subject in the second volume 

 of the 'Illustrations*,' abandons the views which are given by 

 Arnott and himself in the ' Prodromus,' and says that E-oxburgh 

 correctly describes the seed as having a simple inverse embryo the 

 length of the ample perisperm. 



In 1853 Lindleyt, following Hamilton, Roxburgh and Wight, 

 ascribes to the tribe BarringtoniecB an embryo in the axis of 

 copious fleshy albumen. 



Griffith's excellent observations on these anomalous seeds were 

 not published till 1854 J, and consist only of the rough notes 

 attached to the drawings ; the subject, though too important to be 

 entirely neglected by so indefatigable an observer, having only 

 come before him cursorily, so that the notes made at one time 

 are somewhat at variance with those at another. His remarks are 

 so important that I shall quote them at length. At page 657 of 

 the work quoted he says, " The seed of Barringtonia is sufficiently 

 remarkable : I imagine the central part represents above the radicle, 

 below the plumule ; and I also imagine that all the part between 

 the lowest scales of the superficies and the radicle is adherent 

 cotyledon. Or is it an instance of an immense radicle and two or 

 several minute cotyledons (represented by scales and an inconspi- 

 cuous plumule) ? To this, which is suggested by the radicular 

 central system being internal, not reaching the superficies, there 

 is a great objection in the irregularity, in number and situation, of 

 the scales. The internality may arise from the cotyledons being 

 produced into basilar auriculae, concealing and enclosing the roots, 

 as occasionally happens. The development only will clear up the 

 point, for even if the scales become enlarged and foliaceous during 

 * p. 19. t. 100. t Veg. Kingd. 754. % Notulse, vol. iv. 



LINK. PKOO.^ — BOTANY. E 



